Three dashboards, like in older shared versions, but with: - multi-ceph cluster support (selects based on $job) - r/w/rw latencies instead of just apply/commit - added rw for throughput and iops